一种基于心理物理模型的用户依从性评估方法及系统

By constructing and comparing psychophysical models and nested models, and analyzing user response patterns, this study solves the problem of compliance assessment in existing training systems, and achieves objective, low-cost, real-time compliance assessment and personalized training control, applicable to visual, cognitive, and neuromodulation training.

CN122398201APending Publication Date: 2026-07-17JIANGSU JUEHUA MEDICAL TECH CO LTD

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
JIANGSU JUEHUA MEDICAL TECH CO LTD
Filing Date
2026-04-22
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing training systems lack effective methods for assessing compliance, making it difficult to accurately determine user behavior patterns, which affects training effectiveness and data reliability. Furthermore, existing peripheral monitoring devices are expensive or subjective reports are unreliable.

Method used

By constructing a psychophysical model and using maximum likelihood estimation and nested model comparison, we can analyze users' response patterns under different stimulus intensities, calculate compliance scores, and conduct objective evaluations in conjunction with auxiliary information.

Benefits of technology

It achieves objective and low-cost assessment of compliance, can identify the difference between random guessing and stable responses, provides real-time feedback and personalized training control, and is applicable to a variety of task types.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122398201A_ABST
    Figure CN122398201A_ABST
Patent Text Reader

Abstract

本发明公开了一种基于心理物理模型的用户依从性评估方法及系统,涉及医疗健康及行为数据分析技术领域。包括:通过采集用户在训练任务中的刺激强度与应答结果,利用心理物理模型构造不含失误参数的简化模型与包含失误参数的完全模型两种嵌套模型,对刺激强度‑正确反应概率关系进行最大似然估计拟合基于似然比检验失误参数对于用户反应数据的贡献程度,生成0~100范围内的依从性评分。本发明不仅能够解决依从性难以准确判断的问题,还能够提升训练系统对用户状态的识别能力,为训练过程控制、个体化干预策略制定以及临床效果评估提供更可靠依据。
Need to check novelty before this filing date? Find Prior Art